The content factory is closing

The answer to what a software company should write has held for nearly twenty years: publish content that answers questions, rank on Google, turn readers into customers. I’ve been circling it all week, working out what to put on three sites of my own. It’s the question every founder with a keyboard hits eventually. What’s actually worth writing?

That answer has a specific origin. It’s HubSpot’s. They coined inbound marketing, wrote the book on it in 2010, and built a $40 billion company proving it worked. Their blog was the growth engine and the proof of concept at once. Every SaaS blog you’ve ever seen is downstream of that playbook. Including, if I’m honest, the guides section I’m planning for my own product.

Which is why what happened to HubSpot’s blog is worth sitting with.

In March 2023 the blog was pulling around 24 million visits a month. By early 2025 it was closer to 6 million. Nearly 5 million monthly visits vanished between November and December 2024 alone. Three quarters of the traffic, gone in under two years, at the company that invented the model.

Two things did the damage. The first is Google’s AI Overviews. Search “what is a marketing qualified lead” and the answer is right there at the top of the page, synthesised. There’s nothing left to click, so the visit that used to feed the funnel never happens.

The second is more instructive. HubSpot ranked for everything. Famous quotes. Resignation letter templates. Cover letter examples. None of it had anything to do with the CRM software they sell, but it drove traffic, so the content factory kept producing it. Then Google started rewarding topical authority, and a site that ranks for everything becomes hard to trust on anything. The junk reportedly dragged down the pages that were genuinely good. HubSpot ended up quietly deleting tens of thousands of posts.

Volume was the strategy. Volume became the liability.

Here’s the part that doesn’t fit the obituary, though. HubSpot isn’t dying. Revenue is growing and they’re profitable. When AI tools answer questions about CRM software, HubSpot is the most recommended brand in its category. The expertise still pays. It just pays in citations now, not clicks.

That distinction is the whole lesson. The old game rewarded coverage: write about everything adjacent to your space and catch the traffic. If the inventor of the playbook can’t make that work with a dedicated content team, a solo builder writing at nights certainly can’t. The new game rewards being the thing worth citing. When someone asks an AI a question in your niche, the model reaches for the sources that know the subject. A site with fifteen expert pieces on one subject now outranks a site with fifteen hundred pieces on everything.

So the product site gets guides on exactly one subject, the one I deal with every day, and nothing outside it. No “famous engineering quotes” post is ever getting written, whatever the keyword tool says. The personal site gets essays like this one, which were never really for Google anyway.

The tempting read is that content marketing is dead. I think that’s backwards. The content that was always a bit cynical is dead, the stuff written to catch a keyword rather than to say something. What’s left is the writing that was worth doing in the first place. The bar went up. Given what the old bar produced, I’m not sure that’s a tragedy.

The content factory is closing. Good riddance. Write the thing only you can write, about the thing you actually do, and let the machines find it.